Research Spotlight

Introduction Recent natural and man-made disastrous events underscore the importance of emergency evacuation planning and management for buildings and facilities (Averill et al., 2005). Although building codes include provisions for egress design, their prescriptive nature poses many limitations in terms of providing effective evacuation for specific buildings, each of which is unique. Human behaviors under emergency situations have been overly simplified in developing the codes and provisions. For instance, in the “Means of Egress” section of the International Building Code (ICBO, 2000), the methods for calculating the number of exits, width of exits, and the distribution of exits are based on the assumption that occupants are spatially distributed evenly and would utilize all available exits and channels to escape in case of an emergency. Many of such assumptions are not realistic. For instance, people often tend to exit a building based on their familiarity with the surrounding, ignoring many available means in favor of the entrances they normally use. In an emergency situation, people often look for cues and support information and follow the crowds. To fully study egress system through real experiments is not feasible not only because of the cost but they also may require exposing people to dangerous environment. Computer simulations can serve as a viable alternative, under the condition that such simulations are able to capture human and social behaviors in emergency situations. Many computational tools for the simulation of emergency evacuation and design of egress are now available. However, most of current tools focus primarily on the modeling of spaces and occupancies but rarely take into consideration human and social behavior. While these computational models may be more accurately simulate evacuation than simply following codes and guidelines, they tend to rely on a priori assumptions that are difficult to evaluate, inconsistent or unrealistic (Fire Protection Engineers 2002; Santos and Aguirre, 2004). The objective of this research is to develop a computational framework that could facilitate incorporating human and social behavioral models for simulating emergency evacuation. The simulated results could potentially provide better insights into the design of egress and to assess the performance of evacuation systems. This note describes a prototype development of a dynamic, Multi-Agent based Simulation System for Egress analysis (MASSEgress) that is able to incorporate individual’s behavior and to model group interactions (such as competitive behavior, herding behavior, etc.).